Eldorado - Repositorium der TU Dortmund

Ressourcen aus und für Forschung, Lehre und Studium

Bei diesem Service handelt es sich um das Institutionelle Repositorium der Technischen Universität Dortmund. Hier werden Ressourcen aus und für Lehre, Studium und Forschung gespeichert, erschlossen und der Öffentlichkeit zugänglich gemacht.

Hauptbereiche in Eldorado

Wählen Sie einen Bereich, um dessen Inhalt anzusehen.

Gerade angezeigt 1 - 5 von 9

Fach- und Universitätsübergreifendes 1318
Fakultäten 13735
Graduate School und Graduiertenkollegs 12
Hochschulverwaltung 1242
Sonderforschungsbereiche 1592

Aktuellste Veröffentlichungen

Incorporating sufficient physical information into artificial neural networks
(2024-02-19) Geuken, Gian-Luca; Mosler, Jörn; Kurzeja, Patrick
The concept of Rao-Blackwellization is employed to improve predictions of artificial neural networks by physical information. The error norm and the proof of improvement are transferred from the original statistical concept to a deterministic one, using sufficient information on physics-based conditions. The proposed strategy is applied to material modeling and illustrated by examples of the identification of a yield function, the identification of driving forces for quasi-brittle damage and rubber experiments. Sufficient physical information is employed, e.g., in the form of invariants, parameters of a minimization problem, isotropy and differentiability. It is proven how intuitive accretion of information can yield improvement if it is physically sufficient, but also how insufficient or superfluous information can cause impairment. Opportunities for the improvement of artificial neural networks are explored in terms of the training data set, the networks’ structure and output filters. Even crude initial predictions are remarkably improved by reducing noise, overfitting and data requirements.
Improvements in charged lepton and photon propagation for the software PROPOSAL
(2024-05-15) Alameddine, Jean-Marco; Albrecht, Johannes; Dembinski, Hans; Gutjahr, Pascal; Kampert, Karl-Heinz; Rhode, Wolfgang; Sackel, Maximilian; Sandrock, Alexander; Soedingrekso, Jan
Accurate particle simulations are essential for the next generation of experiments in astroparticle physics. The Monte Carlo simulation library PROPOSAL is a flexible tool to efficiently propagate high-energy leptons and photons through large volumes of media, for example in the context of underground observatories. It is written as a C++ library, including a Python interface. In this paper, the most recent updates of PROPOSAL are described, including the addition of electron, positron, and photon propagation, for which new interaction types have been implemented. This allows the usage of PROPOSAL to simulate electromagnetic particle cascades, for example in the context of air shower simulations. The precision of the propagation has been improved by including rare interaction processes, new photonuclear parametrizations, deflections in stochastic interactions, and the possibility of propagating in inhomogeneous density distributions. Additional technical improvements regarding the interpolation routine and the propagation algorithm are described.
Hydrophobic interactions described using hetero-segmented PC-SAFT
(2024-04-09) Rother, Marius; Sadowski, Gabriele
The hydrophobic effect plays a central role in aqueous systems containing molecules with hydrophobic moieties. Despite the relevance of this effect for chemical processes or pharmaceutical applications, modeling remains challenging even using molecular-based models such as different versions of the Statistical Associating Fluid Theory. This work shows the feasibility of hetero-segmented PC-SAFT for this purpose. This model was used as a group contribution method to build molecules from different functional groups, namely from CH2, CH3, and CH2OH groups. Saturated-liquid densities and vapor pressures of n-alkanes and n-alcohols as well as binary mixtures of these molecules were accurately described by this approach. The description of aqueous mixtures containing n-alkanes and n-alcohols was improved compared to state-of-the-art modeling by explicitly accounting for intermolecular interactions resulting in the hydrophobic effect. The so-obtained framework describes the two liquid−liquid equilibrium phases in mixtures of n-alkanes or n-alcohols with water equally well using a single set of transferable parameters. The model was also validated for vapor–liquid equilibria, solid–liquid equilibria, infinite-dilution properties, as well as octanol/water partition coefficients and showed quantitative agreement with experimental data.
Impact of the spindle number on the material transport and mixing during planetary roller melt granulation
(2024-04-01) Lang, Tom; Bartsch, Jens
In comparison to the established twin-screw machines, the application of a planetary roller granulator for continuous operation of melt granulation is a promising alternative based on the unique process concept. An initial study focused on the material transport during processing as a key driver for the overall performance. Hereby, the impact of direct process parameters on the residence time distribution was the main objective. These investigations are complemented in this study by considering the free processing volume, which is defined by the number of planetary spindles applied within a module. The impact on the processing conditions is evaluated with respect to the process setting in terms of feed rate and rotation speed. The results highlight the potential of altering the underlying transport function in planetary roller melt granulation (PRMG) via the investigated direct process and equipment parameters. The impact of the feed rate is lower in comparison, while a higher rotation speed as well as a higher free processing volume promote material mixing. Moreover, a normalization of the determined residence time distribution (RTD) model data was feasible with respect to the process settings and the number of applied planetary spindles in the processing zone. This demonstrates the key role of the free processing volume in the fundamental mechanisms of material transport and mixing during PRMG.
Generating fully developed flow profiles for simulations of viscoelastic fluids described by differential models
(2026-02) Westervoß, Patrick; Damanik, Hogenrich; Mierka, Otto; Ouazzi, Abderrahim; Turek, Stefan
In this paper, an iterative method is developed to compute the flow profiles of a viscoelastic fluid without solvent part described by various differential models in case of a two-dimensional fully developed channel flow. Therefore, the tensorial formulation of the models for the conformation tensor is written as a nonlinear one dimensional system of equations for the unknown components of the conformation tensor, the pressure and the velocity field. On the basis of an initial guess for the pressure drop, the unknown quantities can be determined iteratively. For sufficiently high relaxation times, the resulting flow profiles show typical viscoelastic behaviour. Furthermore, corresponding two-dimensional Finite Element simulations of a Poiseuille flow through a plane channel are used to perform a validation of the determined flow profiles for the Giesekus model.